That's not correct brad. The oatmeal look on top of the melt would more than likely be Sb, not tin.
.
I'm not so sure that it's always antimony trying to creep out of solution that causes that. If I cast 2/3 of a 20 lb pot and throw all the sprues back in as I go with no fluxing or reducing, I get about three heaping teaspoons of clumpy, grainy stuff accumulated on top. A simple wax reduction puts it all back in, which only tells me for certain that it is some sort of oxide accumulation, likely mostly tin-rich "skins" of culled bullets and sprues. When what I *think* might be antimony separating (no sprues put back in, usually casting at low pot temp with alloy like Hardball), it usually takes a load of sawdust to make it go back into solution.
On the subject of dross composition, I know some tests have been done on intentionally generated pot dross and IIRC it had the same content as the alloy from which it came. I don't believe those reports, and here's why: Tin oxidizes on the surface of molten ternary alloy
instantly upon exposure to air. It is this very mechanism that causes tin to "reduce the surface tension" of the alloy and make it flow/cast better. Tin oxide skin is more flexible and thin than antimony or lead oxide skin and lets the melted metal flow like a lava field. Ever notice "tin tails" hanging from your bottom pour spout? That's the tin skin at work. So it follows that if you have a mostly tin or tin/antimony skin forming on the metal surface (and it takes a redox reaction to convert it, not heat), that "dross" will be disproportionately rich in tin compared to the base alloy and you will deplete tin by skimming the pot. Considering tin is usually only present in a few percent, it doesn't take much loss to change the alloy characteristics significantly during the course of a single-pot casting session. Probably not a big deal, but throwing back sprues as you go (I do this sometimes, as a volume expedient) can cause issues that may or may not need to be corrected by a little fluxing, stirring, and reduction.
Type metal vats had to be maintained with some sort of enrichment alloy (high in tin and antimony) to make up for the losses through skimming. I have in my possession a box of monotype letters which mostly test between about eight and 12 on the BHN scale. Obviously they were cast from various depleted alloys.
For what it's worth, I did my own testing of wheel weight plus 2% tin skimmings years ago by reducing the skimmings to a small, clean ingot and comparing specific gravity to that of the base metal. I don't remember the exact number, but I'm pretty sure the skimmings came in around 8.5 grams per CC using a graduated cylinder and a powder scale, draw your own conclusions as to their composition.
It's easy to get wrapped around the proverbial axle on these things, I tend to stick with the Bass Ackwards philosophy and go with what works for what I'm doing, which isn't always exactly the same technique for every casting session.